Agricultural Monitoring and Crop Prediction System with Machine Learning

• Admin login and dashboard
• User management with create, edit, view, activate, deactivate, and delete options
• Farm management for all registered users
• Dataset upload, preview, archive, restore, and training selection
• Data preprocessing with missing value and duplicate inspection
• Machine learning model training for crop recommendation, yield estimation, and risk prediction
• Default model selection for crop, yield, and risk modules
• Model metrics view and trained model management
• Global crop monitoring record management
• Soil, yield, and risk record management
• Prediction record management with update and delete options
• Parameter range management for agricultural input validation
• Feedback management with status update, resolve, reopen, and delete
• Notification and notice management
• Analytics and aggregated reporting dashboard
• CSV and PDF export reports
  • Database backup workflow

• Farmer signup and login system
• User dashboard with farm overview and activity summary
• Profile management with password, security question, contact details, and profile image
• Farm creation, editing, and deletion
• Crop monitoring entry for N, P, K, temperature, humidity, rainfall, soil moisture, pH, season, crop, and yield
• Data validation using admin-defined parameter ranges
• View active training dataset
• Trigger machine learning model training from the user portal
• Crop prediction based on soil nutrients and climate data
• Soil health monitoring with rule-based condition labels
• Yield estimation using trained regression model
• Agricultural risk classification using trained ML model
• Monitoring history management
• Prediction history view
• Personal report generation with chart views
• Download personal CSV and PDF reports
• Feedback submission to administrators

• Built with Flask 3 and Python 3
• Uses SQLAlchemy ORM with SQLite database by default
• Supports local machine learning model training with scikit-learn
• Random Forest classifier and regressor integration
• Pickle model storage under instance/models/
• Automatic creation of required folders and database on first launch
• Default admin account generation
• Default parameter range generation
• Dataset fallback using data/crop_recommendation.csv
• Kaggle dataset compatible crop recommendation workflow
• Extended support for yield and risk columns
• Report generation using ReportLab and Matplotlib
• CSV cleaning and export support
• Suitable for academic, portfolio, internship, and final year project use

1. Clone or download the project.

2. Create a virtual environment:

   python -m venv .venv
   .\.venv\Scripts\Activate.ps1

3. Install dependencies:

   pip install -r requirements.txt

4. Run the Flask application:

   python app.py

5. Open in browser:

   http://127.0.0.1:5000

Optional environment variables

• SECRET_KEY for Flask session security
• DATABASE_URL for custom database connection
• ADMIN_PASSWORD to set initial admin password before first database creation

Seed demo data

python seed_data.py

Administrator

• Username: admin
• Password: admin123

Demo Farmer Users

• Generated after running:

  python seed_data.py

• Default password for seeded farmers: farmer123